The Influence of Alcohol on the Base Excess Parameter in Trauma Patients

• Published in: Shock (Augusta, Ga.) Vol. 56; no. 5; p. 727
• Main Authors: Leiblein, Maximilian, Sturm, Ramona, Franz, Niklas, Mühlenfeld, Nils, Relja, Borna, Lefering, Rolf, Marzi, Ingo, Wagner, Nils
• Format: Journal Article
• Language: English
• Published: United States 01.11.2021
• Subjects: Adult; Aged; Alcoholic Intoxication - blood; Ethanol - blood; Female; Humans; Male; Middle Aged; Prognosis; Retrospective Studies; Wounds and Injuries - blood; Wounds and Injuries - mortality
• ISSN: 1540-0514
• DOI: 10.1097/SHK.0000000000001782

The base excess (BE) parameter can be used as an indicator of mortality. However, study results on the influence of alcohol on the validity of BE as a prognostic parameter in alcohol-intoxicated patients are controversial. Thus, this study examined the hypothesis: An increasing blood alcohol level reduces the prognostic value of the BE parameter on mortality. In a retrospective analysis of the multicenter database of the TraumaRegister DGU, patients from 2015 to 2017 were grouped depending on their blood alcohol level (BAL) into a BAL+ and BAL- group. The hypothesis was verified using logistic regression with an assumed significance level of 1% (P < 0.01). Eleven thousand eight hundred eighty-nine patients were included; 9,472 patients in the BAL- group and 2,417 patients in the BAL+ group. Analysis of the BE showed lower values in the BAL+ group (BAL-: -1.8 ± 4.4 mmol/L vs. BAL+: -3.4 ± 4.6 mmol/L). There is a trend toward lower BE levels when BAL increases. Assuming a linear relationship, then BE decreases by 0.6 points per mille alcohol (95% CI: 0.5-0.7; P < 0.001). The mortality rate was significantly lower in the BAL+ group (BAL-: 11.1% vs. BAL+: 7.9%). The logistic regression analysis showed a significant beneficial influence of BAL+ on the mortality rate (OR 0.706, 95% CI 0.530-0.941, P = 0.018). To analyze whether a low BE (≤-6 mmol/L) has different prognostic effects in patients with and without alcohol, logistic regression models were calculated. However, the effect of BE ≤ -6 mmol/L was similar in both models (regression coefficients in BAL-/+ patients: 0.379/0.393). The data demonstrate an existing influence of alcohol on the BE parameter; however, this does not negatively affect the BE as a prognostic parameter at a threshold of ≤ -6 mmol/L.